The Economic Times, 4 April, 2011
It is never easy counting secretive, solitary predators. For decades, foresters studied pugmarks and usually counted more tigers per tiger. Then, in 2002, Project Tiger (now National Tiger Conservation Authority) and Wildlife Institute of India began replacing the human error-prone pugmark census method with a scientific estimation protocol. It was a landmark initiative.
Nine years later, however, India’s tiger numbers remain equally suspect. So far, more than Rs 22 crore has been spent during two all-India estimation drives, in 2006-07 and 2010-11, to scientifically evaluate the status of the tiger. And yet all the government churned out were a few gospel figures for media consumption.
When the subjective pugmark count method was junked, the promise was of moving away from banal number games towards effective monitoring. Yet, the first all-India tiger estimation report said “these population estimates have high variances, but since these estimates are not to be used for monitoring trends… they should suffice the need for converting a relevant ecological index to a more comprehensible concept of numbers.” Numbers make headlines, more so when spiced up by technical jargons.
Sure enough, when minister for environment and forests Jairam Ramesh proudly announced the gain of 225 tigers last week, it made happy headlines. Accounting for the Sunderbans figures (70) that were not available in 2008, the minister preened, the population gain was a healthy 295.
But the lack of due scientific rigour was soon evident when on March 31, three days after Ramesh had charmed the media, WII rushed to correct its own presentation. An email from a WII scientist to NTCA admitted that “there has been a mistake in the computation of the standard error for the tiger numbers for the state of Maharashtra”, with a request for updating the MoEF website. The Press Information Bureau website, however, is yet to drop the incorrect figures.
To be fair, the WII-NTCA presentation – Status of Tigers in India, 2011 -- the one Ramesh appropriately put up on his ministry’s website, does quote biologists Richard Hutto and Jock Young: “Any monitoring program is a compromise between science and logistic constraints.” While the presentation stops at that cryptic disclaimer, the ministry’s own records do not.
Back in 2006, an international team of experts led by John Seidensticker from the department of conservation biology at Smithsonian’s National Zoological Park in Washington, DC did a peer review of the new estimation method. In its report, the team questioned the feasibility of the exercise given that more than 40,000 forest units would have to be sampled, adding that new method, too, relied on the “integrity of the primary data collectors, data compilers and their supervisors.” The words strangely echoed the NTCA’s own justification for discarding the old pugmark count method.
The peer review also warned that the genetic methods proposed in the census were not “fully developed for this application” and that there were not even enough GPS (global positioning system) sets to map out the terrain.
For the record, the new method breaks down the estimation process in three phases. Phase one involves data collection (signs of tiger presence, prey abundance and canopy status) across the tiger landscapes. In the second phase, satellite data is used to assess conditions of habitat. Phase three requires camera-traps to be set up in selected pockets for capturing tiger images to identify the number of individuals in a sampled area. Next, the camera-trap data is extrapolated to arrive at numbers for entire landscapes.
The observations in the peer review on the challenges of logistics and integrity made the phases one and two look suspect during the first all-India estimation. Little has changed since. What is worse, the phase three of the latest count was compromised by too many malfunctioning camera-traps.
WII purchased around 500 Moultrie camera-traps for the second all-India estimation, out of which 300-odd malfunctioned. The official stand remains that the manufacturer replaced the faulty sets. Sources in the field, however, report a different picture.
In Corbett, around 60 camera traps were installed in two phases (pre and post-monsoon) in 2010. Within days of installation, most cameras reported an activation lag: when an animal passed by, the camera would take several seconds to recover from the sleep mode and click an empty frame or only the animal’s hind portion. Even if the faulty cameras were replaced after monsoon, areas surveyed in the pre-monsoon phase with faulty sets were not covered again.
In Maharashtra’s Tadoba-Andhari tiger reserve, around 60 camera sets malfunctioned when the pre-monsoon camera trapping began last year. While a WII field team returned with replacements after monsoon, it barely spent a month in the field. In some reserves, camera-traps reached only in late October.
Across the world, top institutes like International Snow Leopard Trust uses Reconyx camera-traps. In India, camera-trapping pioneer Dr Ullas Karanth has used analog Trailmaster and later Deercam camera-traps. Among the cheaper brands, Bushnell is believed to be the most versatile. For reasons best known to WII bigwigs, they chose to rely on Moultrie camera-trap and its slow reaction. Maybe, our sarkari scientists expected tigers to stand still and be counted in national interest.
But even this piece of equipment did not make it to many non-reserve areas. Not to mention the tiger-bearing forests outside Protected Areas (PAs). For example, no camera-trap was set up in Tadoba’s Kolsa range, let alone the areas outside the reserve notorious for human-tiger conflict.
WII used about 100 cameras to cover just 120 sq km of Nagarjunasagar-Srisailam tiger reserve in Andhra Pradesh. Only seven tigers were identified in that area. Based on these seven tigers in just 120 sq km, somehow the estimate extrapolated a figure of 60 for a 2342 sq km-area of this reserve.
If data was compromised, the analysis was suspect, too. On the extrapolation process, the 2006 peer review cautioned that “there is also no detailed write-up of the technical analysis, explicitly identifying the analytical techniques to be used in each phase of the framework.’’ As mentioned earlier, the 2008 report did accept “high variances” in estimation. In simple terms, it means different results at different estimation attempts, resulting in unreliability. Nevertheless, those unreliable estimates were further extrapolated to larger areas for arriving at “all-India” figures.
Former WII biologist K Yoganand had pointed out in 2008 how the official count only referred to standard errors of estimated densities and population sizes while offering a range of 1165–1657 tigers. He said the report did not use the appropriate confidence intervals which, if standardised, would have stretched the limits to roughly 900-1900 tigers, an embarrassingly unreliable range.
Surprisingly, the estimation process has not been made public in the last nine years. Only one scientific paper -- Can the abundance of tigers be assessed from their signs? (2011) -- has been published and that too explains only a part of the method.
Clearly, after nine years and more than Rs 22 crore spent, the promise of robust science has not materialised. Instead of quadrennial rituals, the ministry should have opted for intensive assessment, using camera-traps and DNA analyses, of all major source tiger populations, allowing for timely management inputs. For example, four years might be a bit too late for realising that a key tiger reserve like Kanha is losing tigers as has been revealed in the latest estimate. Given that NGOs like Aranyak (Kaziranga) or WWF-India (Ramnagar) did a good job of estimate, the ministry could rope in more organisations for logistics and manpower.
While the number game continues, it is important to note that the new protocol does not allow comparison of subsequent results. The 2008 report noted that “these estimates are not to be used for monitoring trends”. Moreover, a number of tiger areas that were left out during the last estimation, have been covered this time. Comparing the common areas assessed on both occasions, the WII-NTCA report modifies the population growth to 12 per cent or 170 tigers. Insiders, however, claim that the 13 new areas that have been added this time account for a total of 288 tigers. This brings down the actual population gain to just seven tigers or less than 0.5 per cent.
But is a there a case for celebration even if we accept the government’s figures unquestioningly? We may have achieved a 12 per cent rise in our tiger population but lost almost 24 per cent of India’s tiger habitat. Do 170 tigers more or less really make a big difference? Carnivores have high turnover and cats are anyway prolific breeders.
Habitat loss or fragmentation, however, is a long term, often irreversible, damage. Given adequate breeding forests, adding 170 tigers to the national population is always a short-term possibility. In comparison, restoring tigers to 21,000 sq km of habitat lost since 2006-07 will be next to impossible. That our conservation efforts, while rustling up feelgood numbers, have actually conceded one-fourth of the tiger’s home in just four years, is the real headline of this census.
